Customer satisfaction,scored the way the standard defines it.
CSAT is the top-box percentage: how many people picked the happy end of the scale, out of everyone who answered. Half the tools online quietly average the raw ratings instead and hand you a different number. Enter your counts and get the real score, the dissatisfied share hiding behind it, and an honest read on where it lands.
Your numbers
How many people picked the top boxes — a 4 or 5 on a 5-point scale (or whatever you count as 'satisfied').
Everyone who answered the satisfaction question. The denominator is responses, not customers or sends.
The verdict
CSAT
80.0%
Dissatisfied share
20.0%
Satisfied responses
320
Total responses
400
Who answered, split out
Against the 80% rule of thumb
CSAT is the top-box percentage, not the average rating. 80% is the line most teams quote as good — but it's a sanity check, not a benchmark. Read it next to who actually answered.
Reverse-solve · hit a target CSAT
Top-box responses needed
340
More to convert
20
At 400 responses, a 85% CSAT needs 340 of them in the top box. You're 20 short — that's 20 currently-dissatisfied responses you'd need to win over (or new top-box responses on top). The lever is the dissatisfied share, not the people already happy.
Sensitivity · top box swings the score
What it takes to move the number
| Change in top-box responses | Satisfied | CSAT |
|---|---|---|
| -20 | 300 | 75.0% |
| -10 | 310 | 77.5% |
| -5 | 315 | 78.8% |
| No change | 320 | 80.0% |
| +5 | 325 | 81.3% |
| +10 | 330 | 82.5% |
| +20 | 340 | 85.0% |
On 400 responses, each top-box response is worth about 0.3% of CSAT. That's why small samples lurch — and why winning back a handful of detractors moves the number more than chasing the people who are already satisfied.
Your move
The score is the easy part. The 20% who weren't satisfied are the brief.
CSAT 80.0% — 320 of 400 responses satisfied (20.0% not).
A number on a dashboard doesn't fix anything — the verbatims behind it do. Send me your CSAT survey, the comments, and where customers actually drop off, and I'll map which dissatisfaction is costing you revenue and what I'd change first. Free, and the read is yours whether we work together or not.
Plain English
CSAT is a top-box percentage, not an average rating.
Customer Satisfaction Score (CSAT) is the share of respondents who rated you at the satisfied end of the scale. On the standard 5-point scale, that's everyone who picked a 4 or a 5, divided by everyone who answered, times 100. A score of 80 means 80% of people who responded were satisfied — it does not mean the average rating was 4.0 out of 5.
That distinction is where most online calculators go wrong. They quietly average the raw ratings (a 4.2/5 becomes 'CSAT 84') and hand you a number that looks the same but isn't. The two can disagree sharply: a survey full of 3s and 5s can have a healthy average and a mediocre top-box CSAT, because the 3s aren't 'satisfied.' This calculator counts top-box responses, which is how the metric is actually defined and reported.
CSAT is best read as a fast, transactional pulse — satisfaction right after an interaction, a purchase, a support ticket. It's cheap to run and easy to act on, but it has a known blind spot: only some people respond, and happy or angry people respond more, so the score can flatter or punish you. Pair the number with who answered, and with the dissatisfied share, before you draw conclusions.
The formula
CSAT = (Satisfied responses ÷ Total responses) × 100
320 people gave you a 4 or 5 out of the 400 who answered → CSAT = 320 ÷ 400 × 100 = 80%. The other 80 responses (the dissatisfied share, 20%) are the part worth reading verbatim — that's where the next improvement is hiding, not in the 320 who are already happy.
How to read your CSAT (general guidance, not a benchmark)
Satisfaction varies wildly by industry, by how you word the scale, and by who bothers to answer, so there's no single authoritative 'good CSAT' number to compare against. The bands below are a rule-of-thumb sanity check — 80% is the line most teams quote as good — not an industry benchmark:
| CSAT range | Read | What it signals |
|---|---|---|
| 0–70% | Underwater | Most surveys hit this when something is actively broken. The dissatisfied share is large enough to drive churn and word-of-mouth damage. Treat it as a fire, not a trend line. |
| 70–80% | Workable | Respectable but below the line most teams aim for. You're keeping the majority happy and leaving a meaningful slice on the table. Worth digging into the detractors. |
| 80–90% | Good | The common benchmark zone. 80% is the number people quote when they say a CSAT is 'good.' You're doing the fundamentals well; the gains now are in the specifics. |
| 90%+ | Excellent | Top-box dominance. Either you're genuinely delighting people or your sample is skewing positive (happy customers answer more). Verify it's real before you celebrate. |
Source: General rule of thumb (not an industry benchmark) — CSAT, the top-box definition · 2025
Your CSAT looks off. What's it actually telling you?
CSAT high, but very few people responded.
Low response rate inflates CSAT — the indifferent middle stays silent and the happy ones answer. Look at response rate alongside the score. A 95% CSAT on a 4% response rate is a vanity number, not a verdict.
CSAT good, but customers still churn.
CSAT measures the moment, not the relationship. Someone can be satisfied with a support reply and still leave over price or fit. Pair it with NPS (loyalty) and your actual retention curve before you relax.
The dissatisfied share is small but loud.
A 15% dissatisfied share is 15 in every 100 telling their friends. Don't let a passing aggregate score bury the verbatims. Read the detractor comments — the fix is almost always written in them.
Score swings hard month to month.
Small samples are volatile. A 20-response survey moves 5 points if one person changes their mind. Set a minimum sample before you trust a trend, and check it against the survey sample-size math.
How to move CSAT up (and keep it honest)
01Ask at the right moment
Trigger the survey right after the interaction, while it's fresh. A satisfaction question sent a week late measures memory, not the experience.
02Read the detractors first
The 1s and 2s carry your roadmap. Every dissatisfied response is a specific, fixable failure someone took the time to describe.
03Close the loop
Reply to negative responses personally. Recovered detractors often become your loudest advocates, and they tell you exactly what nearly lost them.
04Raise response rate, not just the score
A representative sample beats a flattering one. One short, well-timed question gets answered; a 12-question form gets abandoned by everyone but the extremes.
05Keep the scale consistent
Don't change the wording or the number of points between surveys. The moment you do, your trend line breaks and old scores stop comparing.
06Segment the score
A blended CSAT hides the truth. Split by plan, channel, or new-vs-returning and the real problem (and the real win) usually jumps out.
07Act visibly on what you hear
When customers see their feedback change something, satisfaction rises on its own. 'You asked, we shipped' is the cheapest CSAT lever there is.
08Don't game the top box
Begging for 5s or excluding unhappy customers from the send corrupts the metric. A true 82 beats a manufactured 96 every time.
The vocabulary
- CSAT
- Customer Satisfaction Score: the percentage of respondents who chose the satisfied (top-box) ratings, out of everyone who answered.
- Top box
- The highest one or two points on the scale (a 4 or 5 on a 5-point scale) — the responses that count as 'satisfied' in the CSAT formula.
- Dissatisfied share
- 100 minus your CSAT: the percentage of respondents who were not satisfied. Often the more actionable half of the result.
- Response rate
- The share of people who answered the survey. A high CSAT on a low response rate is easy to over-trust.
- NPS
- Net Promoter Score: a loyalty metric (would you recommend us?) measured on a 0–10 scale. Measures the relationship where CSAT measures the moment.
- CES
- Customer Effort Score: how hard it was to get something done. A strong predictor of loyalty, and a useful companion to CSAT.
CSAT questions, straight answers
CSAT = (satisfied responses ÷ total responses) × 100. Count everyone who picked the top boxes — a 4 or 5 on a 5-point scale — and divide by everyone who answered the question. 320 satisfied out of 400 responses is a CSAT of 80%. It's a top-box percentage, not the average of the raw ratings.
Keep going
A calculator tells you what. A call tells you what to do about it.
Send me the account behind these numbers. I'll tell you straight where the money's leaking and what I'd fix first — free, and you keep it whether you hire me or not.